Innovation is often the result of novel combinations and interdisciplinary collaboration. Biochemistry, medicine, and technology are three fields of science that have been closely intertwined for decades. At Aqsens Health biochemistry and machine learning have morphed in a novel fashion, creating new and exciting possibilities in screening, diagnostics and in health care.
The partnership between biochemistry and medicine
Biochemistry and medicine are two very closely connected disciplines that benefit from a mutually cooperative relationship. A deep understanding of a disease always requires an understanding of the biochemical processes involved in the disease mechanism.
Biochemistry is also crucial in the development of new diagnostic tools, which is why it is also at the heart of Aqsens Health’s research. Aqsens Health’s E-TRF–method is the innovation of Professor Hänninen, who has led its development from the beginning.
Professor Pekka Hänninen’s academic and professional career has been an interesting combination of engineering, computer science, and biology. Currently he is a professor of Medical Physics and Engineering at the University of Turku. Professor Hänninen’s interdisciplinary background has provided him with a unique perspective on the technological developments and applications of biochemistry in healthcare, and also on combining scientific expertise from different fields.
“Nowadays all new discoveries come from a combination of different fields of science. Interdisciplinary cooperation is very important in making new innovations. It’s a fact,” Professor Hänninen reflects.
In terms of biochemistry, AI, and healthcare, combining the already powerful field of biochemistry with technological advancements has the potential to significantly improve both the efficiency and accuracy of screening and diagnostics. Machine learning (ML) especially is widely applied in different ways in healthcare, and ML combined with novel biochemistry is an important part of the development of Aqsens Health’s screening tests.
What is machine learning?
ML algorithms are present everywhere in our daily lives – from social media applications to online banking and autopilot on commercial flights. And the number of different ML applications is only growing. Machine learning is usually defined as a branch of AI. Computer scientist and ML pioneer Tom Mitchell defines it as “the study of computer algorithms that improve automatically through experience”. Essentially, ML is one of the key building blocks of an AI.
Machine learning applications and algorithms are based on learning through processing data. By learning from existing data, a ML algorithm is able to improve its performance and accuracy without explicitly being programmed to do so.
“There are several different types of learning when it comes to AI. The type of learning we are currently utilizing is called supervised learning. Learning based on knowledge. The other type of learning connected to AI is called unsupervised learning. In that case, the algorithm looks for different patterns in large data sets, possibly finding some new connections,” Professor Hänninen explains.
Strengthening E-TRF with machine learning
Aqsens Health’s use of machine learning algorithms is slightly different to the more common approaches that use ML.
The E-TRF method measures the luminescence of a sample. Because of our proprietary chemical and biochemical modulators, the luminescence signals between healthy control samples and disease samples are different. Machine learning plays a role in recognizing these disease signatures.
“In our method, machine learning is connected to holistic chemistry, which allows us to interpret our results in more detail. That is what we are doing, teaching the machine learning algorithm to recognise tiny disease patterns in biological samples, which enables us to examine even the smallest details in the sample,” Professor Hänninen summarizes.
Aqsens Health is currently taking major steps in the research laboratory in using biopanned phages to increase E-TRF’s specificity to detect the severity of certain diseases, like lethal prostate cancer. Later on in the process, machine learning will have an even more important role.
“I believe that this combination of biochemistry and machine learning will play an important role in years to come. Integrating our technologies and algorithms to larger healthcare systems will significantly improve the efficiency of the healthcare decision making process,” Professor Hänninen concludes.
As early detection and prevention become key factors in maintaining and improving public health actions all over the world, it is important that the healthcare industry utilizes the innovations of the new age efficiently.